I also feel that last statement is true: not that they left because of the need to build another wallet, but because they had a more general plan for multi-chain architecture and knew they would need a particular kind of multi-chain wallet to support it, and all their pending contributions to the wallet space are being applied there.
So Flint is being designed for more general use cases than the first emerging Cardano wallets, including multi-blockchain compatibility; soon it will be able to talk to EVMs like Milkomeda without having to use MetaMask as an auxiliary interface for Ethereum addresses 
…
see:
The Milkomeda Foundation has been in discussions with dcSpark to push for adding native Milkomeda C1 support inside of Flint Wallet
From what I know of the designers I think they may simply not have posted yet about security principles because they generally incorporate best practices about pretty much everything. So Flint is also my own tentative answer to the question in this other thread (I think my own statement here was referenced in the OP of this thread):